Molecular imaging approaches for metabolic and physiologic imaging of tumors have become important for treatment planning and response monitoring. However, the relationship between the physiologic and metabolic aspects of tumors is not fully understood. Here, we developed new hyperpolarized MRI and electron paramagnetic resonance imaging procedures that allow more direct assessment of tumor glycolysis and oxygenation status quantitatively. We investigated the spatial relationship between hypoxia, glucose uptake, and glycolysis in three human pancreatic ductal adenocarcinoma tumor xenografts with differing physiologic and metabolic characteristics. At the bulk tumor level, there was a strong positive correlation between 18F-FDG-PET and lactate production, while pO2 was inversely related to lactate production and 18F-2-fluoro-2-deoxy-D-glucose (18F-FDG) uptake. However, metabolism was not uniform throughout the tumors, and the whole tumor results masked different localizations that became apparent while imaging. 18F-FDG uptake negatively correlated with pO2 in the center of the tumor and positively correlated with pO2 on the periphery. In contrast to pO2 and 18F-FDG uptake, lactate dehydrogenase activity was distributed relatively evenly throughout the tumor. The heterogeneity revealed by each measure suggests a multimodal molecular imaging approach can improve tumor characterization, potentially leading to better prognostics in cancer treatment.

Significance:

Novel multimodal molecular imaging techniques reveal the potential of three interrelated imaging biomarkers to profile the tumor microenvironment and interrelationships of hypoxia, glucose uptake, and glycolysis.

Cancer cells acquire aberrant biochemical pathways to support their uncontrolled growth. The abnormal energetic demands of malignant cells differ significantly from normal cells to support metabolic activities, which generate building blocks for further growth in anabolic processes (1). Tumors adapt to their microenvironment by altering their metabolism in response to hypoxia, nutrient starvation, immune surveillance, and other stressors (2–4). These alterations, known as metabolic reprogramming, provide a clear biochemical phenotype for detection and grading and may guide potential treatment options (5). The unique microenvironment of tumors offers opportunities for targeted therapeutic strategies. Several inhibitors that preferentially target these bioenergetic pathways are now in preclinical and clinical trials (6). As development of these inhibitors progresses, metabolic imaging may become increasingly relevant for pharmacodynamic validation of drug action beyond its current role in developing diagnostic and prognostic information (7, 8).

The current standard for cancer diagnosis, staging, and treatment management by metabolic imaging is 18F-2-fluoro-2-deoxy-D-glucose (18F-FDG)-PET, which monitors the uptake and subsequent phosphorylation of the nonmetabolizable glucose analogue fluorodeoxyglucose. Because the upregulation of glucose transport is a common feature of many cancers and is frequently used to monitor treatment prognosis, FDG-PET has found widespread use in cancer diagnosis and staging (5). As FDG-PET probes only the first steps of glucose uptake and phosphorylation and is insensitive to the exact chemical nature of the tracer, other methods that analyze the enzymatic reactions downstream of entry in metabolism are advantageous. 13C magnetic resonance spectroscopy has been suggested as an alternative but suffers from low sensitivity. To bring the signal of the tracer to a level that can be detected by MRI, it is necessary to use the process of hyperpolarization to transfer the larger spin polarization of the unpaired electron on a paramagnetic molecule to the 13C-labeled tracer under nonequilibrium conditions.

Hyperpolarized MRI has several theoretical advantages over FDG-PET for metabolic imaging (9, 10). Hyperpolarized MRI can be particularly advantageous in organs such as the brain that have a naturally high glucose uptake, which hinders utilizing 18F-FDG-PET imaging approaches due to the large background from normal tissue (5). Hyperpolarized pyruvate in particular has proved useful as the pyruvate to lactate kinetics of conversion measures flux through the critical switching point from glycolysis to the tricarboxylic acid cycle (11).

Given that FDG-PET is already in widespread clinical use and has proven effective in multiple clinical trials, the question naturally arises about what new information hyperpolarized MRI can bring. Although FDG-PET and pyruvate hyperpolarized MRI in theory measure two distinct metabolic processes, glucose uptake and the Warburg effect, they share common regulation points through transcriptional control of both glucose transporter 1 (GLUT1) and lactate dehydrogenase A (LDHA) by hypoxia-inducible factor 1 (HIF1) and cMyc. Both processes are known to exist in the same tumor microenvironment and hence the delivery of a bolus of pyruvate or 2[18F]fluoro-2-deoxy-D-glucose (FDG) is affected similarly in both cases by the tumor vasculature and tumor interstitial pressure.

Prior studies have probed the relationship between FDG-PET uptake and hyperpolarized MRI. In voxel-wise comparisons of rat hepatocellular carcinomas, Menzel and colleagues found no significant correlation between lactate to pyruvate ratios in hyperpolarized MRI and standardized uptake values (SUV) in FDG-PET (9). On the other hand, a strong positive voxel-wise and patient-wise correlation was found between FDG uptake and lactate/pyruvate in sarcomas but not carcinomas in canine cancer models (10). Differences in vasculature along with differences in intrinsic metabolism were speculated to be responsible for these differences. To provide better insight into these discrepancies, we compared FDG-PET and hyperpolarized MRI measurements in three well-characterized pancreatic ductal adenocarcinoma (PDAC) mouse xenograft models along with pO2 imaging through electron paramagnetic resonance (EPR). We show that the successful comparison of multimodal modalities allows us to monitor tumor physiology in a complemental manner, and provides insights into detailed assessment of in vivo tumor metabolism and microenvironment.

### Animal studies

All of the animal experiments were carried out in compliance with the Guide for the Care and Use of Laboratory Animal Resources, and experimental protocols were approved by the Animal Care and Use Committee, NCI (NCI-CCR-ACUC; ref. 12). The human pancreatic tumor cell lines, Hs776t, and SU.86.86 cells were obtained from Threshold Pharmaceuticals, and MiaPaCa2 cells were purchased from the ATCC. All cell lines were authenticated by IDEXX RADIL utilizing a panel of microsatellite markers, and tested Mycoplasma negative by Frederick National Laboratory for Cancer Research. Hs766t, MiaPaCa2, and SU.86.86 pancreatic tumor–bearing mice (n = 9 each) were created by injecting 3 × 105 cells subcutaneously into the right hind legs of athymic nude mice.

### Hyperpolarized 13C MRI

Hyperpolarized 13C MRI experiments were performed on a 3T MRI Scanner (MR Solutions Inc.) using a 17-mm diameter home-built 1H/13C coil. A 96 mmol/L hyperpolarized [1-13C] pyruvate solution from a Hypersense DNP Polarizer (Oxford Instruments) was administrated via a tail veil cannula (12 μL/g body weight). 13C two-dimensional spectroscopic images were acquired with a 32 × 32 mm2 field of view in a 8-mm axial slice through the tumor.

### 18F-FDG-PET

Tumor-bearing mice were injected with 100 μCi of 18F-FDG in PBS via a tail veil cannula under anesthesia. Sixty minutes after 18F-FDG administration, a PET scan was conducted using a BioPET/CT (Bioscan Inc.) under anesthesia with 1.5% isoflurane with a nominal resolution of 0.375 mm × 0.375 mm × 0.375 mm.

### EPR imaging

A 300 MHz pulsed EPR imaging scanner with a tailor designed parallel coil was used for oxygen imaging using OX063 as the paramagnetic tracer to obtain pO2 maps with a resolution of 0.4375 mm × 0.4375 mm × 0.4375 mm.

### Molecular imaging establishes a relationship between tumor oxygenation, glucose uptake, and glycolysis at the bulk tumor level in PDAC xenografts

pO2, FDG uptake, and lactate/pyruvate ratios for MiaPaCa2, Hs766t, and SU.86.86 xenografts were correlated at the whole-tumor level (Fig. 1AC). All three cell lines were derived from PDAC tumors and have a largely similar genetic background (13), but differ strongly in the tumor microenvironment. Similar to previous reports, we found a strong global inverse relationship between tumor oxygenation and lactate/pyruvate, a measurement of LDH activity with a saturating bolus of pyruvate (14). At the whole-tumor level, FDG uptake was strongly correlated with lactate/pyruvate and inversely correlated with pO2. Out of the three xenografts, SU.86.86 most closely resembled normal tissue in all measures, in-line with its higher degree of vascularization and differentiation. The cardinal ordering for each measure (SU.86.86>MiaPaCa2>Hs766t for pO2 and the reverse Hs766t>MiaPaCa2>SU.86.86 for FDG uptake and lactate/pyruvate) was in overall agreement with the corresponding biomarkers (Fig. 2A and B). GLUT1 transporter and, to a lesser degree, hexokinase-2 were expressed more highly in Hs766t, which showed elevated FDG uptake and retention relative to the other two PDAC xenografts. Similarly, the xenograft with the highest median pO2 value, SU.86.86, also showed the highest expression levels of the angiogenesis biomarker CD31, while the highest expression levels of HIF1 were found in Hs766t, the cell line with the lowest median pO2 value. The ordering of LDH enzyme expression matched the ordering of lactate/pyruvate, as expected by the relationship between lactate/pyruvate ratios and pyruvate to lactate enzymatic flux (15).

Figure 1.

Multimodal imaging identifies the relationships between tumor oxygenation, glucose uptake, and glycolysis at the whole-tumor level. A, pO2 maps derived from EPR oximetry, 13C two-dimensional spectroscopic images, and 18F-FDG PET images for the cell lines indicated (top to bottom). B, Median pO2, lactate/pyruvate (lac to pyr),18F-FDG SUV uptake, and hypoxic fraction (<10 mmHg) on a whole tumor basis for the indicated PDAC cell lines. C, Correlation between the indicated measures when considered on a whole tumor basis (n = 9; error bars, SD). *, P < 0.05, t test.

Figure 1.

Multimodal imaging identifies the relationships between tumor oxygenation, glucose uptake, and glycolysis at the whole-tumor level. A, pO2 maps derived from EPR oximetry, 13C two-dimensional spectroscopic images, and 18F-FDG PET images for the cell lines indicated (top to bottom). B, Median pO2, lactate/pyruvate (lac to pyr),18F-FDG SUV uptake, and hypoxic fraction (<10 mmHg) on a whole tumor basis for the indicated PDAC cell lines. C, Correlation between the indicated measures when considered on a whole tumor basis (n = 9; error bars, SD). *, P < 0.05, t test.

Close modal
Figure 2.

Protein expression levels of key proteins associated with metabolism support multimodal imaging results. A, Expression levels of CD31, HIF1, LDHA, GLUT1, hexokinase-2, and MCT1 from tumor extracts. B, These expression levels were determined by Western blot analysis. Error bars, SD (n = 3). *, P < 0.05, t test.

Figure 2.

Protein expression levels of key proteins associated with metabolism support multimodal imaging results. A, Expression levels of CD31, HIF1, LDHA, GLUT1, hexokinase-2, and MCT1 from tumor extracts. B, These expression levels were determined by Western blot analysis. Error bars, SD (n = 3). *, P < 0.05, t test.

Close modal

Strong correlations existed between bulk measurements of pO2, FDG uptake, and lactate/pyruvate in xenografts from different PDAC cell lines when values were summed across the entire tumor (Fig. 1C). While these correlations are useful in establishing a correspondence between different metabolic properties of the tumor in bulk, the advantage of imaging studies is that they can address the inherent heterogeneity of tumors to reveal how metabolic processes are linked to the tumor environment, which varies considerably within the tumor.

### Multimodal imaging reveals the intrinsic heterogeneity of the physiologic and metabolic aspects of tumors, linked to the tumor environment

Representative 0.46-mm slices from EPR pO2 imaging of MiaPaCa2, SU.86.86, and Hs766t xenografts were selected (Fig. 3A). In contrast to hyperpolarized MRI and FDG-PET images, which have a variegated appearance, indicating substantial tumor heterogeneity on at least the millimeter scale, the only noticeable feature of the pO2 images was a contiguous hypoxic volume near the center of the tumor, whose size was reflective of the known angiogenic potential of the cell lines from which they were derived (Fig. 3A). Partially due to this relative smoothness, there was a modest correlation on a voxel-by-voxel basis between pO2 levels and FDG-PET (Fig. 3B). On larger distance scales there was a rough inverse correlation between pO2 and FDG uptake; metabolic activity in general was highest in the interior of the tumor where pO2 was at its lowest point (Fig. 3C). This was also observed in coregistered images, which overlaid the hypoxic fraction, the radio-insensitive region of the tumor where pO2 was less than 10 mmHg with the FDG uptake (Fig. 3A). Although the relationship is not precise, a significant fraction of highly metabolically active cells lay in the well-oxygenated region outside the hypoxic fraction that is predominantly in the tumor core. Correlations between pO2 values and FDG uptake calculated from a voxel-by-voxel comparison for each slice indicated that FDG uptake was negatively correlated with pO2 in the center of the tumor and positively correlated with pO2 on the exterior regions.

Figure 3.

Registered FDG-PET and oximetry images highlight the inherent tumor heterogeneity within the tumors. A, Registered slices from FDG-PET and EPR oximetry from the central region of the tumor. Each 64 × 64 slices was 0.375 mm in thickness; the slice number is indicated in white at the bottom left corner of each image. The hypoxic fraction (pO2 < 10 mmHg) is outlined in blue on the FDG images. B, Left, mean FDG uptake values for the hypoxic (red) and nonhypoxic fraction (green) of xenografts from different PDAC cell lines for the slices defined in A. Slice 32 corresponds to the center of the tumor. Right, correlation between pO2 values and FDG uptake calculated from a voxel by voxel comparison for each slice from registered images. FDG-PET uptake was negatively correlated with pO2 in the center of the tumor and positively correlated with pO2 on the surface. C, Mean intensities from a 10 × 10 voxel patch at the center of each tumor.

Figure 3.

Registered FDG-PET and oximetry images highlight the inherent tumor heterogeneity within the tumors. A, Registered slices from FDG-PET and EPR oximetry from the central region of the tumor. Each 64 × 64 slices was 0.375 mm in thickness; the slice number is indicated in white at the bottom left corner of each image. The hypoxic fraction (pO2 < 10 mmHg) is outlined in blue on the FDG images. B, Left, mean FDG uptake values for the hypoxic (red) and nonhypoxic fraction (green) of xenografts from different PDAC cell lines for the slices defined in A. Slice 32 corresponds to the center of the tumor. Right, correlation between pO2 values and FDG uptake calculated from a voxel by voxel comparison for each slice from registered images. FDG-PET uptake was negatively correlated with pO2 in the center of the tumor and positively correlated with pO2 on the surface. C, Mean intensities from a 10 × 10 voxel patch at the center of each tumor.

Close modal

Quantitatively, there was a modest but consistent pixel-wise correlation between both the pyruvate and lactate signals from hyperpolarized MRI and FDG-PET (Fig. 4A and B). The interpretation of this relationship is complicated by the fact that lactate production reflects both the uptake of pyruvate into the tissue and its metabolic conversion; the lactate signal is highly correlated with pyruvate as the variance in pyruvate transport and bolus delivery is much larger than the difference in LDH activity. Because FDG and pyruvate uptake are dependent on the same vascular system for delivery, they can be expected to be correlated to some degree. Using the lactate/pyruvate ratios reduces this dependence, for in simplified models with first order kinetics, the ratio is strictly equal to the rate of the conversion of pyruvate to lactate (16). Using lactate/pyruvate to adjust for uneven uptake of pyruvate, LDH activity was distributed mostly uniformly throughout the tumor with higher activity in the tumor interior. On a quantitative level, this was seen by comparing each modality to a distribution that was completely uniform throughout the tumor, using the Kullback–Leibler divergence metric (KL) to measure how far each modality was from a uniform reference image. A KL of zero in this case indicated an image that was completely uniform; higher values indicated heterogeneity. Lactate/pyruvate was notably closer to uniform than either lactate or pyruvate alone or FDG uptake or pO2 levels (Fig. 4C). While hypoxia and FDG-PET uptake appeared to be partially linked at least on a gross-anatomic level, the higher uniformity of lactate/pyruvate suggested LDH activity was driven by other factors.

Figure 4.

Registered images show consistent pixel-wise correlations quantitatively. A, Comparison of a 8-mm center slice for each of the imaging modalities indicated as T2-weighted MRI, SUV from FDG-PET, and pyruvate (Pyr), lactate (Lac), and the pyruvate/lactate ratio (Lac/Pyr) from 13C hyperpolarized MRI. The FDG-PET and EPR images were resampled to 14 × 14 and an 8-mm thickness to match the resolution of the hyperpolarized MRI image. The lactate/pyruvate ratio was set to zero outside the boundary of the tumor as defined by the FDG-PET image. B, Correlation between methods calculated by a voxel by voxel comparison of registered images. C, The Kullback–Leibler divergence (⁠KL\ = \sum_{i = 0}^n {P_i}{\rm{log}}( {\frac{{{P_i}}}{{{Q_i}}}} )$⁠, where {P_i}$ represents pixels from the image and {Q_i}\ $the corresponding pixel from the reference image) of the 14 × 14, 8-mm center slices of each modality in A compared with a tumor that was completely uniform with the same mean value. Images formed from the lactate/pyruvate ratio from hyperpolarized MRI were significantly more uniform than images from any other measure. For each comparison A–C, all images are rescaled to the resolution of the HP image and each value is calculated on a per animal basis. Error bars, SD (n = 3). Figure 4. Registered images show consistent pixel-wise correlations quantitatively. A, Comparison of a 8-mm center slice for each of the imaging modalities indicated as T2-weighted MRI, SUV from FDG-PET, and pyruvate (Pyr), lactate (Lac), and the pyruvate/lactate ratio (Lac/Pyr) from 13C hyperpolarized MRI. The FDG-PET and EPR images were resampled to 14 × 14 and an 8-mm thickness to match the resolution of the hyperpolarized MRI image. The lactate/pyruvate ratio was set to zero outside the boundary of the tumor as defined by the FDG-PET image. B, Correlation between methods calculated by a voxel by voxel comparison of registered images. C, The Kullback–Leibler divergence (⁠KL\ = \sum_{i = 0}^n {P_i}{\rm{log}}( {\frac{{{P_i}}}{{{Q_i}}}} )$⁠, where {P_i}$represents pixels from the image and {Q_i}\$the corresponding pixel from the reference image) of the 14 × 14, 8-mm center slices of each modality in A compared with a tumor that was completely uniform with the same mean value. Images formed from the lactate/pyruvate ratio from hyperpolarized MRI were significantly more uniform than images from any other measure. For each comparison A–C, all images are rescaled to the resolution of the HP image and each value is calculated on a per animal basis. Error bars, SD (n = 3).

Close modal

Availability of new clinical imaging modalities naturally raises the question of what comparative advantage the new methods have against the existing. FDG-PET is well-established clinically and is an essential technique for tumor diagnosis and treatment planning. However, FDG-PET only probes the first stage of glycolysis, glucose uptake, and is unable to probe deeper into the metabolic and physiologic differences that distinguish tumors from normal tissues and tumor from each other. These metabolic and physiologic differences in the tumor microenvironment can potentially be targeted in multiple ways. Hypoxia in particular is an important marker for predicting radiation sensitivity and can serve as guide for directing treatment but has not yet found widespread usage due to the difficulties in imaging hypoxic regions (17). However, while glucose uptake is not a direct marker for hypoxia, they share a pathway through HIF activation of the GLUT1 transporter (15). Similarly, the regulation of GLUT1 and hexokinase-2 is tied to the regulation of other metabolic genes through Ras and other oncogenes (18).

On the basis of this linked regulation, it has been proposed that glucose uptake through FDG-PET may serve as a surrogate for hypoxia and general metabolism. Previous preclinical in vivo imaging studies have suggested that a correlation between hypoxic regions and glucose metabolism exists with minor locoregional differences in some cases (19). The data for this hypothesis is however limited and conflicting (11), and in the case of hypoxia, complicated by the difficulties of accurately measuring pO2 (19). We found each measure is largely independent with a modest correspondence when evaluated on a voxel by voxel basis, although a considerable correlation exists between whole-tumor measurements. The FDG-PET and pO2 maps showed a similar geometric distribution within the tumor, possibly due to a higher metabolic demand under hypoxic conditions. The positive and negative correlations between pO2 and FDG uptake seen in Fig. 3B and C are likely driven by differing rates of oxygen consumption caused by differences in metabolism. In the tumor core where the correlation is negative, the tumor is likely quiescent and oxidative phosphorylation serves as the main energy source as the demand for metabolic building blocks produced by glycolysis is low. High oxygen consumption in these regions creates a state of consumptive hypoxia. While poor perfusion would also explain the low pO2 observed in the tumor core, it would also limit FDG uptake as O2 and glucose are expected to perfuse approximately equally. In the more invasive regions, at the rim of the tumor, rapid growth stimulates glycolysis to supply metabolic demand for proteins and nucleotides, leading to a positive correlation between FDG uptake and pO2, as the effects of consumptive hypoxia are limited. It should be noted, however, that this study was limited to imaging of a series of less differentiated xenografts from cell lines of a single cancer type and caution should be applied in generalizing this model. Higher resolutions studies of a more diverse set of cell lines, as well additional measurements of perfusion and oxygen consumption, would be helpful in establishing the validity of the model for tumor metabolism in general (Supplementary Fig. S1; refs. 20, 21).

The lactate/pyruvate ratio, which serves as a surrogate for LDH activity, is distributed relatively uniformly throughout the tumor in these models and largely orthogonal to both FDG-PET and pO2, with a modest correlation to each measure as shown in Fig. 4B and C. Hyperpolarized MRI therefore offers additional information that cannot be inferred from either FDG-PET or measures of hypoxia and may prove a useful clinical adjunct to either, although the limited resolution in human subjects remains obstacle. As clinical imaging has been transforming conventional medicine, mapping metabolic activity noninvasively can potentially contribute to better prognostics, further tumor characterization, and earlier response monitoring in cancer treatment.

No potential conflicts of interest were disclosed.

The content of this article does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.

Conception and design: K. Yamamoto, J.B. Mitchell, M.C. Krishna

Development of methodology: K. Yamamoto, J.R. Brender, M.C. Krishna

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): K. Yamamoto, T. Seki, S. Kishimoto, N. Oshima, R. Choudhuri, E.M. Jagoda

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): K. Yamamoto, J.R. Brender, S.S. Adler, E.M. Jagoda, P.L. Choyke

Writing, review, and/or revision of the manuscript: K. Yamamoto, J.R. Brender, P.L. Choyke, J.B. Mitchell, M.C. Krishna

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): N. Oshima, E.M. Jagoda, K. Saito, N. Devasahayam

Study supervision: M.C. Krishna

This research was supported by the Intramural Research Program of the NCI, NIH. This project has been funded in whole or in part with federal funds from the NCI, NIH, under contract no. HHSN261200800001E.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
DeBerardinis
RJ
,
Lum
JJ
,
Hatzivassiliou
G
,
Thompson
CB
.
The biology of cancer: metabolic reprogramming fuels cell growth and proliferation
.
Cell Metab
2008
;
7
:
11
20
.
2.
Ward
PS
,
Thompson
CB
.
Metabolic reprogramming: a cancer hallmark even Warburg did not anticipate
.
Cancer Cell
2012
;
21
:
297
308
.
3.
Biancur
DE
,
Paulo
JA
,
Malachowska
B
,
Quiles Del Rey
M
,
Sousa
CM
,
Wang
X
, et al
Compensatory metabolic networks in pancreatic cancers upon perturbation of glutamine metabolism
.
Nat Commun
2017
;
8
:
15965
.
4.
Dewhirst
MW
,
Secomb
TW
.
Transport of drugs from blood vessels to tumour tissue
.
Nat Rev Cancer
2017
;
17
:
738
50
.
5.
Gillies
RJ
,
Beyer
T
.
PET and MRI: is the whole greater than the sum of its parts?
Cancer Res
2016
;
76
:
6163
6
.
6.
Rai
G
,
Brimacombe
KR
,
Mott
BT
,
Urban
DJ
,
Hu
X
,
Yang
SM
, et al
Discovery and optimization of potent, cell-active pyrazole-based inhibitors of lactate dehydrogenase (LDH)
.
J Med Chem
2017
;
60
:
9184
204
.
7.
Vander Heiden
MG
,
DeBerardinis
RJ
.
Understanding the intersections between metabolism and cancer biology
.
Cell
2017
;
168
:
657
69
.
8.
Wolpaw
AJ
,
Dang
CV
.
Exploiting metabolic vulnerabilities of cancer with precision and accuracy
.
Trends Cell Biol
2018
;
28
:
201
12
.
9.
Menzel
MI
,
Farrell
EV
,
Janich
MA
,
Khegai
O
,
Wiesinger
F
,
Nekolla
S
, et al
Multimodal assessment of in vivo metabolism with hyperpolarized [1-C-13]MR spectroscopy and F-18-FDG PET imaging in hepatocellular carcinoma tumor-bearing rats
.
J Nucl Med
2013
;
54
:
1113
9
.
10.
Hansen
AE
,
Gutte
H
,
Holst
P
,
Johannesen
HH
,
Rahbek
S
,
Clemmensen
AE
, et al
Combined hyperpolarized C-13-pyruvate MRS and F-18-FDG PET (hyperPET) estimates of glycolysis in canine cancer patients
.
2018
;
103
:
6
12
.
11.
Gray
LR
,
Tompkins
SC
,
Taylor
EB
.
Regulation of pyruvate metabolism and human disease
.
Cell Mol Life Sci
2014
;
71
:
2577
604
.
12.
National Research Council (US) Committee for the Update of the Guide for the Care and Use of Laboratory Animals
.
Guide for the care and use of laboratory animals
.
Washington, D.C
:
Institute of Laboratory Animal Resources (US)
;
2011
.
13.
Deer
EL
,
Gonzalez-Hernandez
J
,
Coursen
JD
,
Shea
JE
,
Ngatia
J
,
Scaife
CL
, et al
Phenotype and genotype of pancreatic cancer cell lines
.
Pancreas
2010
;
39
:
425
35
.
14.
Wojtkowiak
JW
,
Cornnell
HC
,
Matsumoto
S
,
Saito
K
,
Takakusagi
Y
,
Dutta
P
, et al
Pyruvate sensitizes pancreatic tumors to hypoxia-activated prodrug TH-302
.
Cancer Metab
2015
;
3
:
2
.
15.
Singh
D
,
Arora
R
,
Kaur
P
,
Singh
B
,
Mannan
R
,
Arora
S
.
Overexpression of hypoxia-inducible factor and metabolic pathways: possible targets of cancer
.
Cell Biosci
2017
;
7
:
62
.
16.
Hill
DK
,
Orton
MR
,
Mariotti
E
,
Boult
JKR
,
Panek
R
,
Jafar
M
, et al
Model free approach to kinetic analysis of real-time hyperpolarized 13C magnetic resonance spectroscopy data
.
PLoS One
2013
;
8
:
e71996
.
17.
Overgaard
J
.
.
J Clin Oncol
2007
;
25
:
4066
74
.
18.
Ying
H
,
Kimmelman
AC
,
Lyssiotis
CA
,
Hua
S
,
Chu
GC
,
Fletcher-Sananikone
E
, et al
Oncogenic Kras maintains pancreatic tumors through regulation of anabolic glucose metabolism
.
Cell
2012
;
149
:
656
70
.
19.
Dierckx
RA
,
de Wiele
CV
.
FDG uptake, a surrogate of tumour hypoxia?
Eur J Nucl Med Mol Imaging
2008
;
35
:
1544
9
.
20.
Dong
T
,
Liu
Z
,
Xuan
Q
,
Wang
Z
,
Ma
W
,
Zhang
Q
.
Tumor LDH-A expression and serum LDH status are two metabolic predictors for triple negative breast cancer brain metastasis
.
Sci Rep
2017
;
7
:
6069
.
21.
Feng
Y
,
Xiong
Y
,
Qiao
T
,
Li
X
,
Jia
L
,
Han
Y
.
Lactate dehydrogenase A: a key player in carcinogenesis and potential target in cancer therapy
.
Cancer Med
2018
;
7
:
6124
36
.